Skip to content
Skills Library
Data & analytics · Building & coding

role-database:vector-databases

/role-database:vector-databases

Deep operational guide for 16 vector databases. Pinecone (serverless, hybrid search), Weaviate (vectorizers, generative search), Milvus/Zilliz (GPU, index types), Qdrant (quantization, filtering), ChromaDB, pgvector (HNSW/IVFFlat), LanceDB, Vespa, Marqo, Turbopuffer. Use when implementing semantic search, RAG pipelines, recommendation engines, or AI/ML embedding storage.

Licence
MIT
Last repository change
24.07.2026

Assessment and security

80/ 100Class: good
Format15 / 25
Frontmatter with name and description, valid name, description up to 1,024 characters, name matching the folder.
Discoverability25 / 25
Description explains when to use the skill, has a reasonable length and uses third person.
Content25 / 35
At least 150 words and at most 500 lines, headings, step lists, examples or code.
Hygiene15 / 15
No placeholders such as TODO, no Windows paths, with a licence.
Riskno signals detected in the scan
Files in this package
6
Scripts
0
HTML comments
0
URLs in SKILL.md
0
Deprecation

No signals detected. Repository changed on 24.07.2026, without references to retired models or APIs.

Licence

MIT, open: use and modify while retaining author attribution.

What it does and when to use it

Deep operational guide for 16 vector databases. Pinecone (serverless, hybrid search), Weaviate (vectorizers, generative search), Milvus/Zilliz (GPU, index types), Qdrant (quantization, filtering), ChromaDB, pgvector (HNSW/IVFFlat), LanceDB, Vespa, Marqo, Turbopuffer. Use when implementing semantic search, RAG pipelines, recommendation engines, or AI/ML embedding storage.

Original author description

Deep operational guide for 16 vector databases. Pinecone (serverless, hybrid search), Weaviate (vectorizers, generative search), Milvus/Zilliz (GPU, index types), Qdrant (quantization, filtering), ChromaDB, pgvector (HNSW/IVFFlat), LanceDB, Vespa, Marqo, Turbopuffer. Use when implementing semantic search, RAG pipelines, recommendation engines, or AI/ML embedding storage.

How it works

  1. When to use this skill
  2. Core Principles
  3. Reference Pointers
Original SKILL.md headings · 26 lines, 204 words
SKILL.md preview
---
name: role-database:vector-databases
description: |
  Deep operational guide for 16 vector databases. Pinecone (serverless, hybrid search), Weaviate (vectorizers, generative search), Milvus/Zilliz (GPU, index types), Qdrant (quantization, filtering), ChromaDB, pgvector (HNSW/IVFFlat), LanceDB, Vespa, Marqo, Turbopuffer. Use when implementing semantic search, RAG pipelines, recommendation engines, or AI/ML embedding storage.
allowed-tools: Read, Grep, Glob, Bash
---

You are a vector database specialist informed by the Software Engineer by RN competency matrix.

## When to use this skill

Load this skill for semantic search, RAG pipeline design, embedding storage, recommendation engines, or any task requiring approximate nearest-neighbor (ANN) search across vectors.

## Core Principles

- Match database to workload: serverless (Pinecone, Turbopuffer) for variable load, self-hosted (Qdrant, Milvus) for control, pgvector when PostgreSQL is already in the stack
- Always index payloads/metadata before filtering — post-filtering on raw vectors kills performance
- Quantization (scalar int8 → 4x, binary → 32x) is the single best cost-reduction lever
- Hybrid search (dense + sparse BM25) consistently outperforms pure vector search in production RAG

## Reference Pointers

Load the relevant reference file for implementation details:

| File | When to load |
|------|-------------|
| `references/pinecone.md` | Pinecone serverless/pod setup, hybrid sparse-dense, namespaces, collections, Inference API |
| `references/weaviate-milvus.md` | Weaviate vectorizers, generative search, multi-tenancy; Milvus index types, GPU, partitions, RRF hybrid |
| `references/qdrant-chroma.md` | Qdrant quantization, payload filtering, snapshots; ChromaDB embedded collections |
| `references/pgvector.md` | pgvector HNSW/IVFFlat indexes, hybrid FTS+vector SQL, performance tuning, batch upsert |
| `references/selection-guide.md` | Database comparison table, embedding model selection, RAG patterns, index tuning, cost optimization |
Files in this package (6)
  • SKILL.md 2 KB
  • references/pgvector.md 3 KB
  • references/pinecone.md 3 KB
  • references/qdrant-chroma.md 4 KB
  • references/selection-guide.md 5 KB
  • references/weaviate-milvus.md 5 KB

Versions

The database has no other skills named role-database:vector-databases.

Database snapshot: 13.11.2026 · package version a5f0596fb78e (SHA-256 digest of the entire package)

Downloads in the last 30 days: 0 · ranking

Report a problem with this skill: harmful content, incorrect descriptions or an author’s removal request.

Similar skills

  • 100vector-databases
    Vector database integration for embeddings and similarity search. Pinecone, Weaviate, Qdrant, ChromaDB, pgvector. Index management, metadata filtering, hybrid search, and production optimization. USE WHEN: user mentions "vector database", "embeddings", "similarity search",…
  • 90vector-database-engineer
    Provides vector database and semantic search patterns for Pinecone, Weaviate, Qdrant, Milvus, and pgvector in RAG and recommendation systems. Use when implementing vector search or when the user mentions vector database, semantic search, embeddings, or similarity search.
  • 90vector-database-engineer
    Design vector storage, indexing, and semantic retrieval with Pinecone, Weaviate, Qdrant, Milvus, or pgvector. Use when implementing RAG, similarity search, or recommendation systems.
  • 95data-vector-databases
    Use this skill when the user says 'vector database', 'vector db', 'pgvector', 'qdrant', 'milvus', 'pinecone', 'weaviate', 'hnsw', 'ivf', 'product quantization', 'ann search', 'similarity search', 'embeddings storage', 'hybrid search'. Chooses a vector store, designs…
  • 100vector-database
    Use for all vector database work — pgvector schema design, Weaviate collection creation, RAG pipeline scaffolding, embedding model selection, index tuning (HNSW vs IVFFlat), and embedding model migration. Triggers on: 'vector search', 'pgvector', 'weaviate', 'embedding', 'RAG…
  • 100vector-db
    Use when operating a vector store as a data layer — choosing or migrating between Pinecone, Qdrant, Weaviate and pgvector; designing a collection or index (distance metric, dimensions, HNSW parameters, named vectors); filtering on metadata; hybrid dense-plus-sparse search; and…

Feedback and reports

Your message goes to Bartosz Sokoliński. An email address is needed only for a reply. See the privacy policy.